The President's New Freedom Commission on Mental Health (2003) proposed that mental health care in the United States be transformed to a consumer oriented recovery model, and in response the Veterans Administration and many State governments have begun reorganizing their programs and committed large sums of money to implementing a recovery model throughout their systems of care. However, this systems change is being driven by social mandate and consensual agreement rather than by empirical support. There is little scientific literature on the nature of recovery or factors that influence it, and recovery oriented treatment models have not been subjected to systematic evaluation. Two factors that have limited empirical study of the recovery construct and recovery-oriented treatment are: a) the absence of a scientifically grounded conceptual model for recovery, and b) the lack of a reliable and valid assessment instrument to measure recovery status. The purpose of this project is to develop a psychometrically sound, practical instrument to assess recovery in veterans with serious mental illness using Bandura's social cognitive theory as a conceptual model for the recovery construct. A sample of 200 veterans with serious mental illness will be recruited and assessed on the Maryland Assessment of Recovery in Serious Mental Illness (MARS) measures of agency and self-efficacy, recovery related constructs, and a battery of symptom, neuropsychological, and other outcome variables. We will also collect information from subjects and/or medical records on a range of demographic, treatment, health, and life circumstance variables. The data will allow us to examine aspects of the construct validity of social cognitive theory as a model for recovery, as well as validity of the MARS. All subjects will be reassessed on the MARS and a subset of instruments from the Baseline assessment after 12-months. Test-retest reliability of the MARS will also be examined on 50 participants by administering the MARS on a second occasion approximately 1-week after the Baseline assessment. The Specific Aims are: 1) to evaluate and refine the draft version of the MARS, 2) evaluate its test-retest reliability and validity, and 3) to examine recovery status and the relationship of recovery to hypothesized mediators and moderators over a 1-year retest interval. .